def genCurSubset(opts, n, size): """ Generates the n-th subset. @type opts: Options @param opts: Options object @type n: Number @param n: Number of the subset @type size: Number @param size: Size of the subset @rtype: (String, String) @return: Tupel with sub- and restset """ dataset =dataSetFilename(opts.training_file, n) sub = subsetFilename(opts.training_file, n) rest = restSetFilename(opts.training_file, n) subset(opts, size, dataset, sub, rest) return (sub,rest)
def genCurSubset(opts, n, size):
"""
Generates the n-th subset.
@type opts: Options
@param opts: Options object
@type n: Number
@param n: Number of the subset
@type size: Number
@param size: Size of the subset
@rtype: (String, String)
@return: Tupel with sub- and restset
"""
dataset = dataSetFilename(opts.training_file, n)
sub = subsetFilename(opts.training_file, n)
rest = restSetFilename(opts.training_file, n)
subset(opts, size, dataset, sub, rest)
return (sub, rest)
def run(self):
"""Thread.run()"""
try:
if (self.opts.action == "gui"):
import svmToolGui
svmToolGui.gui = svmToolGui.svmToolGui(self.opts)
svmToolGui.gui.run()
else:
log("SVM Tool {0}".format(version))
log("Using libSVM: {0}".format(self.opts.libsvm_path))
log("\tsvm-scale: {0}".format(self.opts.svm_scale_path))
log("\tsvm-train: {0}".format(self.opts.svm_train_path))
log("\tsvm-predict: {0}".format(self.opts.svm_predict_path))
log("\tsubset.py: {0}".format(self.opts.subset_py))
if (self.opts.gnuplot_path != None):
log("Using gnuplot: {0}".format(self.opts.gnuplot_path))
if (self.opts.training_file != None):
log("Training file: {0}".format(self.opts.training_file))
if (self.opts.test_file != None):
log("Test file: {0}".format(self.opts.test_file))
if (self.opts.model_file != None):
log("Model file: {0}".format(self.opts.model_file))
if (self.opts.output_file != None):
log("Output file: {0}".format(self.opts.output_file))
if (self.opts.action == "crossValidation"):
log("Starting cross validation")
log("C-Range (start,stop, step): {0}".format(
self.opts.c_range))
log("Gamma-Range (start,stop, step): {0}".format(
self.opts.g_range))
log("Division depth: {0}".format(self.opts.division_depth))
log("Division factor: {0}".format(
self.opts.division_factor))
log("Minimum subset size: {0}".format(
self.opts.subset_size))
log("Worker count: {0}".format(self.opts.worker))
if (self.opts.additional_arguments != None
and self.opts.additional_arguments != ""):
log("Additional arguements: {0}".format(
self.opts.additional_arguments))
crossValidate(self.opts)
elif (self.opts.action == "train"):
log("Trainig SVM")
log("C value: {0}".format(self.opts.c_value))
log("Gamma value: {0}".format(self.opts.g_value))
if (self.opts.additional_arguments != ""):
log("Additional arguements: {0}".format(
self.opts.additional_arguments))
train(self.opts)
elif (self.opts.action == "test"):
log("Testing SVM")
if (self.opts.additional_arguments != ""):
log("Additional arguements: {0}".format(
self.opts.additional_arguments))
predict(self.opts)
elif (self.opts.action == "merge"):
log("Merging files")
mergeFiles(self.opts.training_file,
self.opts.output_file + ".data")
mergeFiles(self.opts.test_file,
self.opts.output_file + ".test")
elif (self.opts.action == "subset"):
log("Generating subset")
log("Subset size: {}".format(self.opts.subset_size))
subset(self.opts, self.opts.subset_size, self.opts.training_file, \
self.opts.output_file+".subset_"+str(self.opts.subset_size), \
self.opts.output_file+".restset")
log("All done")
except Exception as e:
log("An exception occured:")
log("\t" + e.message)
log("Here's the stack:\n")
(t, v, s) = sys.exc_info()
traceback.print_tb(s)
return self
def run(self):
"""Thread.run()"""
try:
if (self.opts.action == "gui"):
import svmToolGui
svmToolGui.gui = svmToolGui.svmToolGui(self.opts)
svmToolGui.gui.run()
else:
log("SVM Tool {0}".format(version))
log("Using libSVM: {0}".format(self.opts.libsvm_path))
log("\tsvm-scale: {0}".format(self.opts.svm_scale_path))
log("\tsvm-train: {0}".format(self.opts.svm_train_path))
log("\tsvm-predict: {0}".format(self.opts.svm_predict_path))
log("\tsubset.py: {0}".format(self.opts.subset_py))
if (self.opts.gnuplot_path != None):
log("Using gnuplot: {0}".format(self.opts.gnuplot_path))
if (self.opts.training_file != None):
log("Training file: {0}".format(self.opts.training_file))
if (self.opts.test_file != None):
log("Test file: {0}".format(self.opts.test_file))
if (self.opts.model_file != None):
log("Model file: {0}".format(self.opts.model_file))
if (self.opts.output_file != None):
log("Output file: {0}".format(self.opts.output_file))
if (self.opts.action == "crossValidation"):
log("Starting cross validation")
log("C-Range (start,stop, step): {0}".format(self.opts.c_range))
log("Gamma-Range (start,stop, step): {0}".format(self.opts.g_range))
log("Division depth: {0}".format(self.opts.division_depth))
log("Division factor: {0}".format(self.opts.division_factor))
log("Minimum subset size: {0}".format(self.opts.subset_size))
log("Worker count: {0}".format(self.opts.worker))
if (self.opts.additional_arguments != None and self.opts.additional_arguments != ""):
log("Additional arguements: {0}".format(self.opts.additional_arguments))
crossValidate(self.opts)
elif (self.opts.action == "train"):
log("Trainig SVM")
log("C value: {0}".format(self.opts.c_value))
log("Gamma value: {0}".format(self.opts.g_value))
if (self.opts.additional_arguments != ""):
log("Additional arguements: {0}".format(self.opts.additional_arguments))
train(self.opts)
elif (self.opts.action == "test"):
log("Testing SVM")
if (self.opts.additional_arguments != ""):
log("Additional arguements: {0}".format(self.opts.additional_arguments))
predict(self.opts)
elif (self.opts.action == "merge"):
log("Merging files")
mergeFiles(self.opts.training_file, self.opts.output_file+".data")
mergeFiles(self.opts.test_file, self.opts.output_file+".test")
elif (self.opts.action == "subset"):
log("Generating subset")
log("Subset size: {}".format(self.opts.subset_size))
subset(self.opts, self.opts.subset_size, self.opts.training_file, \
self.opts.output_file+".subset_"+str(self.opts.subset_size), \
self.opts.output_file+".restset");
log("All done")
except Exception as e:
log("An exception occured:")
log("\t" + e.message)
log("Here's the stack:\n")
(t,v,s) = sys.exc_info()
traceback.print_tb(s)
return self
